Thursday, September 30, 2004

Alex Cora Does What He Does Best...

...and Dave Ross and Hee Seop Choi both come up with big extra base hits in extra innings? That's good stuff. Yesterday, when the game was tied after 8 with Beltre due up in the ninth, I started to research to find the last time a team won three games in a row on "walkoff" hits, but didn't get very far by the time the Dodgers lost. Well, the Dodgers did it three times in this series. I'm not up for seeing how common that is, but I'm considering venturing a guess on how likely it is that over two consecutive series against the same team one team will win three games in the bottom of the ninth or the bottom of an extra inning and two other games in which the team trailed by five runs. Yeah, needing this many chances to beat the Rockies doesn't bode very well for the Dodgers, but it's still fun anyway. Now the Dodgers just need to win one game against the Giants to clinch, and even if swept they'll make the playoffs if the Astros lose a game. If the Astros and Giants both come away with sweeps this weekend, the Dodgers then play a tiebreaker game against in San Francisco for the division and the loser plays the Astros on Tuesday for the wild card. Plus, Jon Weisman's prediction over at Dodger Thoughts a while back that either Ross or Mayne would deliver a big hit at some point in the season became even more accurate.

(note: I updated the playoff scenario Friday morning after I actually looked at the standings-- apparently remembering every team's record off the top of my head can get a little hairy.)

Around the Game In Eighty Sentences

I don't think anybody's freaking out about this (it didn't even warrant mention in today's Under the Knife), but I wanted to beat whoever would to the punch:
Asked how Gagne was injured, trainer Stan Johnston told the Los Angeles Times, "Overuse, or it might have been warming up too quick."
Gagne's innings by month since becoming a full-time reliever, with MIG being multi-inning games (games where he was credited with 1.1 or more IP):




IP Pitches MIG
IP Pitches MIG
IP Pitches MIG
April 13 178 1
14.1 179 2
10.2 170 3
May 14.1 231 2
12.2 195 1
11.1 153 1
June 14.1 190 3
12.1 184 0
11 156 3
July 10 140 0
12.2 200 3
15 227 4
August 16.2 270 3
17.1 252 3
17 276 4
September 14 229 3
13 179 4
14.1 232 5
Total 82.1 1238 12
82.1 1189 13
79.1 1214 20

So Gagne has been used slightly more in this stretch run than the two past ones, but not by much. His 735 pitches from July to August is easily his most thrown in a three-month period, and his 46.1 innings in that period is his highest total in any three month period. Likewise, his August-September 2004 slightly edges out August-September in both IP and pitches. However, none of those differences are significant enough for it to be predictable that his usage pattern since the all-star break would constitute overuse. If you think the one extra multi-innings appearance per month is at fault, you'll have to excuse me if I look at you cross-eyed. While it's possible that the way the Dodgers have used Gagne is responsible for his whatever-he's-got (I don't want to call it something and then look stupid when tomorrow's Under the Knife comes out), the Dodgers' decision to mildly increase his usage was the right one since a) he's one of the best three relievers in baseball, b) there's no reason to expect him not to be able to handle the workload, and c) it could turn out he's good for 105-115 IP a year (which I wouldn't put out of the realm of possibility even after this incident) in which case it would behoove the Dodgers to find that out.


You know what? I'm just gonna come right out and say it. I tip-toed around it in my entry on the Cubs' OBP, and it's totally Pollyanna, but I can't help myself. This should be the Cubs' base lineup:
When Grudzie and/or Bako play, everyone else slides up and they come in at the end. If Nomar were actually healthy he'd be higher up, of course. But, I mean, does Dusty Baker realize when he's filling out the lineup card that a. Corey Patterson doesn't get on base too often and b. they're still supposed to play offense after the first inning? Patterson's stolen bases are gravy in tie games in the 8th and 9th, but of limited utility in the first inning.

Alright, file that one under useless (and probably under "Shut up, Everyone Else Has Written the Same Thing, You Just Didn't Notice").


When was the last time I wrote something about the A's? It's been a couple weeks, and I really don't want to pile on. Have you heard the pitching hasn't been tearing it up recently? Bobby Crosby made things better today, but they've been pretty bleak.

The good news is that all they have to do to make the playoffs is win this series, and it's in Oakland. To continue my paean to remembering the import of home-field advantage started earlier today (which, for those counting, will count in my 80 sentence quota), check out Oakland's splits. The Oakland Coliseum (and its various nomenclatural incarnations) has typically played as a pitcher's park but actually has had a hitter-favoring park factor in 2002 and 2004. I can't think of much that makes it unique; lots of foul ground, sure, but that's about it. Meanwhile, the A's have just smoked the competition at home (numbers through Wednesday:

A's in Oakland: 50-27, .272 GPA, 5.13 runs/game, .247 opponents' GPA, 4.02 ERA
A's elswhere: 39-42, .258 GPA, 4.79 runs/game, .255 opponents' GPA, 4.29 ERA


And you know what else drives me crazy that I remembered once again while researching that? Why are there no readily-available home and away fielding splits? I don't know if anyone keeps that data, but it seems to me it should be quite useful to look at team home and away fielding splits and even player home and away fielding splits. You know what? I'd like to make the first effort (well, maybe somebody else has done this, but I couldn't find them using Google-- actually, the Google search turned up this very site). Unfortunately, I can't find any team pitching splits online that include plate appearances or AB, SH, and SF, which means I can't even calculate BABIP, much less DER. But I can find opponents Caught Stealing numbers, so I can make a close-enough formula that doesn't account for double-plays and other baserunning outs: QuasiDER = 1 - ((H - HR)/(3IP + H - HR - SO - CS)). Here goes (%H is how much better the team's DER was at home):



Ana 772 94 752 598 23 0.707
678 76 668.3 533 21 0.707 0
Ari 724 106 715 585 23 0.713
733 88 694 546 26 0.701 1.8
Atl 748 84 748 552 19 0.716
695 64 676 457 10 0.712 0.5
Bal 745 81 700 508 18 0.703
700 73 710.3 553 19 0.713 -1
Bos 732 77 741 584 17 0.712
663 78 675.3 522 14 0.718 -1
ChC 667 88 718 641 38 0.718
655 73 708.3 665 16 0.713 0.8
ChW 806 127 739 531 26 0.71
671 94 658.3 462 21 0.721 -2
Cin 739 125 716 514 17 0.725
829 107 688.3 457 12 0.689 5.3
Cle 788 94 749 584 17 0.703
738 104 689 513 22 0.707 -1
Col 906 110 733 488 20 0.68
693 80 667.7 437 11 0.717 -5
Det 764 94 718 481 27 0.711
740 95 677.7 479 13 0.705 0.8
Fla 678 83 734 603 18 0.727
669 77 672 493 20 0.717 1.3
Hou 665 89 710 664 20 0.715
724 83 706 589 19 0.702 1.9
KC 774 84 689 445 16 0.699
826 118 695.3 422 17 0.699 0
LA 651 85 704 535 15 0.734
695 86 711.3 509 25 0.724 1.3
Mil 719 83 753 597 19 0.721
677 79 655 470 10 0.713 1.1
Min 705 79 731 597 20 0.716
732 81 706.7 500 22 0.711 0.7
Mon 721 86 735 520 25 0.723
730 102 686 491 16 0.712 1.6
NYM 720 64 720 493 22 0.715
707 89 702 460 17 0.725 -1
NYY 746 87 737 570 22 0.711
756 92 672.7 467 10 0.699 1.7
Oak 696 81 718 505 22 0.726
734 80 717.3 505 26 0.713 1.8
Phi 719 111 718 541 13 0.725
738 99 708.7 505 13 0.716 1.3
Pit 721 57 731 533 23 0.711
712 88 671 517 15 0.704 1.1
SD 719 74 728 545 16 0.716
706 105 679 501 9 0.718 -0
SF 787 77 744 547 15 0.702
675 81 677.7 448 9 0.726 -3
Sea 698 104 729 543 19 0.732
765 102 695 464 19 0.707 3.5
StL 658 73 706 515 12 0.731
687 92 711.7 487 15 0.733 -0
TB 680 92 715 471 19 0.738
733 94 658 426 11 0.706 4.5
Tex 780 95 733 493 19 0.711
716 85 671.7 460 20 0.709 0.4
Tor 785 92 708 490 26 0.699
689 84 678 442 15 0.723 -3
Avg 734 89 726 542 20 0.715
716 88 686.3 493 16 0.712 0.4

Well, this doesn't make much of a case for there being an overwhelming case for homefield advantage helping defense, although it shows a slight advantage. Most teams have been better at home so it could be said that the Rockies' numbers skew the results, but that's misleading since two thirds of these teams play games there too which should even out the Coors effect. In fact, this shows that, for this season, anyway, much bigger aspects of the home-field advantage are home runs and strike outs: home teams have given up 4.47% fewer home runs per inning and have struck out 4.12% more batters per inning. In case you're wondering, they've also caught their opponents stealing 16.24% more often, and I'm not sure what to make of that figure. Actually, if we add caught stealing back into the equation, the home team's defensive advantage goes up by one half of one tenth of a percent. The home runs part makes some sense since home teams are more likely, I guess, to adjust their approach based on the relative ease of hitting home runs to various parts of the field and because of rosters constructed to exploit the park's effects. But the strikeouts? I'm not willing to venture any guesses there.

Park Factors and Homefield Advantage

Aaron Gleeman has an article about park factors up at the Hardball Times. In it, he mentions the phenomenon of SBC Park going from a heavy pitcher's park in recent years to a good hitter's park this year. I think there's actually a fairly simple explanation for some of this.

The Giants have an overwhelming home field advantage due to their quirky ballpark, or at least announcers constantly refer to this as if it were true. There's wind, funky angles, all that jazz. I'm betting it's largely true. This means that the Giants are more likely to get outs at home and their opponents are more likely to screw up defensively at SBC. In the past few years, the Giants have had good offenses but better pitching and excellent defense. They had a .751 defensive efficiency ration in 2002 and a .755 DER in 2003; this year they've got a .693 DER. Moreover, the Giants have pretty terrible pitching this year but have had excellent bullpens in each of their previous Pac Bell years and typically very good rotations (especially in 2002 when Rueter had some deity's assistance in compiling a 3.23 ERA). I don't have any significant evidence to support this claim, but I think there's some logic to the argument that the Giants' poor defense and pitching has eliminated much of their home-field advantage in preventing runs. Moreover, the Giants are a better offensive team this year than in years past and perhaps are actually playing to the park's nuances better. The Giants have 168 doubles and 27 triples in 82 games at SBC and 137 2B and 6 3B in 75 games on the road. Their opponents line is 163/25 at SBC, 139/21 away from it. Maybe the Giants' coaching staff figured out that, you know, it might just be a good idea to hit a lot of long fly balls because crazy stuff's going to happen out there. SBC still plays as a pitcher's park for home runs, so there's certainly some strength behind this argument. I think that overall the park factor is likely to fluctuate based on the Giants' particular strengths (as well as the coaching staff's ability to take advantage of the park), in addition, of course, to yearly fluctuation/sample size issues.

Any lessons to be learned from this? Probably not, but I wonder if maybe visiting teams should try out some of this hit it high and far business until the Giants can put out some decent outfielders. This is pretty much anecdotal evidence, but in their ten games at SBC this year, the Dodgers had 12 doubles and one triple while the Giants had 24 doubles and two triples. The pitching and talent levels might not be so much the issue here, as the Dodgers took five of those games (although they were outscored in the process). As Rich Lederer reminded us by channeling Bill James brilliantly once again, homefield advantage can be huge; even more huge might be the ability of some teams to get a thorough understanding of the effects of their opponents' parks and adjust their player's play and personnel decisions accordingly. Which segues right back into... weeklong obsession with the Bobby Cox era. Since 1991, the Braves home winning percentage is .628 and their road winning percentage is .590. Seems pretty fair for a team with their talent. But consider that the this year (sorry, I need to sleep so I can't compile the data for the whole 1991-2004 period) home teams have a .540 win percentage. If that distribution for team won-loss records was uniform, you could find a team's expected road WPCT by multiplying .460 by (home WPCT/.540). That gives the Braves an expected road WPCT of .535 from 1991-2004 and .515 in 2004. Their actual road WPCT has been .590 for that whole period and .577 in 2004. Now, the Braves haven't had as extreme of a park factor all that time as some other teams in the league so perhaps their home/road split shouldn't be expected to be quite so large, and it can be argued that they squandered some potential home-field advantage by building around pitching while they played in slug-heavy Fulton County Stadium. But I have an inkling that part of what Bobby Cox' squad does so well is advanced scouting and quality coaching that reduce their opponents' home-field advantage.

Song of the Day: "I Might Be Wrong," Radiohead

Wednesday, September 29, 2004

More Worthless Manager of the Year Talk

I know I said yesterday that I don't care, and maybe this is just because I don't read enough of the pundits who talk about this business (I've only heard Bobby Cox and Tony La Russa mentioned), but shouldn't Phil Garner be NL Manager of the Year, hands down?

Astros under Williams: 44-44
Astros under Garner: 45-26

I know, I know, this is kind of silly. You'll say Carlos Beltran and I'll say he's given Garner 5 wins above replacement while he gave Williams a half and Richard Hidalgo gave Williams two and a half. That's three wins. And the Astros have had more luck keeping runs from scoring since the break, getting a 4.01 ERA with a .261/.331/.416 line versus 4.14 and .255/.320/.418 before the break (although maybe that's related to the manager). I don't think much of all this manager of the year business, but if Jack McKeon gets it for overseeing Dontrelle Willis and Miguel Cabrera shouldn't Garner get it?

Besides, I'm not sold this is Bobby Cox' greatest showing. Sure, the Braves weren't being picked by anyone, but that had to do more the perception that the Phillies were good and that J.D. Drew would a) be injured and b) play to his previous performance level. Cox should be in the Hall of Fame (as should Leo Mazzone), and I won't rule out the possibility that he's responsible for the incredible seasons from Drew and Estrada. But the Braves "surprise" is equal parts Bad Phillies, Good Drew, Good Estrada, Good Jaret Wright, and Good John Thomson, and if forced to choose I'd attribute those to Bowa, health, fortune, Mazzone, and Mazzone.

You know what? All this manager of the year talk is making me feel unclean. Somebody stop me. Too bad I don't have a reader mailbag to fall back on at times like this (though you're welcome to help me start one).

Another Around the Game in Eighty Sentences, Half-Length Version

Song of the Day: "Tears at the Birthday Party," Elvis Costello and Burt Bacharach

Tuesday, September 28, 2004

Around the Game in Eighty Sentences, Half-Length Version

I must confess that I'm not too interested in who wins the Manager of the Year award in each league, and I really don't have much of an opinion about who should win it. That having been said, there's a lot of popular discourse proclaiming the invincibility of Buck Showalter's candidacy and I thought it would be interesting to examine his case.

Showalter has avoided wasting outs pretty well. His team has only had 23 sacrifices, and only Toronto and Boston have used fewer. The Rangers have probably attempted too many stolen bases this year, as their 36 times caught stealing rank tied for 15th in the major leagues. The Rangers also haven't given up too many opportunities to get outs defensively as they've issued only 29 intentional walks, only five more than league-leading St. Louis.

It's quite unclear how much credit should go to Showalter for the improvement of the pitching staff's performance, and I'd be willing to bet that Orel Hersheiser is probably more deserving of it than Showalter. Similarly, the maintained level of offense despite trading A-Rod for Soriano probably doesn't have much to do with Showalter. If there's a nit to be picked here, it's probably that the Rangers don't take a lot of walks. They've drawn only 483 which is only 23rd best in baseball. Obviously, the personnel the Rangers have has something to do with it; nobody's going to get Alfonso Soriano to walk like Kevin Youkilis. But this is one of those managerial grey areas where a manager might have a significant effect on performance. In Showalter's nine years as a manager, his team has never had an Isolated Discipline (OBP - BA) more than .015 better than league average and half of his teams have had an IsoD lower than league average. Considering the talent he's presided over, that could (just could, nothing more) be a warning sign of sorts.

The most obvious way a manager affects a team is through lineup construction. Showalter gets a mixed review here. One thing that he does well is not giving up outs at the #2 spot. He's been hitting Hank Blalock there almost all season long, a very wise move considering that many managers think there's some constitutional requirement to trot out a weak-hitting middle-infielder there. He's also typically played his better on-base folk in the leadoff spot-- typically Michael Young but occasionally the elder Young, though that should underscore the team's underwhelming overall on-base percentage. His decisions at the 3 and 4 spots in the lineup, however, have been more questionable. Alfonso Soriano has hit #3 all season long while Brad Fullmer hit cleanup while he was still around. That Fullmer hit .233/.310/.442 this year and remained in front of Mark Texeira in the lineup for three months is pretty inexcusable, as Teixeira leads the Rangers in slugging and trails only Eric Young in OBP. It's true that Texas has a fair share of left-handed hitters which means that LOOGY issues should come into play in lineup construction, but Showalter's solution is sub-optimal. Showalter probably should have gone with this as his regular lineup once it became clear that Teixeira had improved his game:

vs. RHP
M. Young, .313/.355/.482
Blalock, .272/.353/.502
Teixeira, .273/.365/.555
Mench, .275/.325/.531
Dellucci, .241/.339/.442
Soriano, .280/.324/.484
Fullmer/Matthews, .233/.310/.433 and .275/.350/.461
Barajas, .248/.275/.451
Nix, .249/.295/.434

vs. LHP
E. Young, .294/.384/.382
Blalock, .272/.353/.502
Teixeira, .273/.365/.555
M. Young, .313/.355/.482
Mench, .275/.325/.531
Soriano, .280/.324/.484
Matthews, .275/.350/.461
Jordan, .227/.280/.372
Barajas, .248/.275/.451

(note: all figures are full season, not pitcher splits; I'm getting lazy on you)

Over his career, Showalter has had issues with lineup construction. He wasted a lot of Paul O'Neill's productivity by having him hit #5 for most of his time in New York, hitting relative out-machine Don Mattingly at #2 or #3 much of that time. As his current use of Blalock would indicate, he's typically not put poor hitters in the #2 spot, and except for 2000 and 1992 he's normally had one of his best 2 or 3 hitters hit second. However, Showalter's been guilty of a tragic case of Womackphilia, with known symptoms including getting an out from one's leadoff hitter 78% of the time over a two-year period. He's also had pretty terrible OBP's from his leadoff men in 1992, 1993, and 2003, so his claim to managerial genius is pretty poor in my book. I gathered a lot of data and made a few charts about all this Showalter lineup construction business, but they're too boring for me to post here; e-mail me if you want them.


I know I covered part of this earlier, but this is getting absurd. The Dodgers took two out of three in a series a week and a half ago in which they trailed all three games by five or more runs. Now, they've taken the first two games of this series in which they trailed in the eighth inning and won in the bottom of the ninth, including five in the ninth to win 5-4 tonight. I couldn't see the game on TV, so I'm weary about what the deal with Milton Bradley was. I'm wondering if my objectivity is strained vis-a-vis Bradley. When Ross Porter was saying what was happening, my reaction was one of outrage at the fans, even before Porter said that someone threw a beer bottle at Bradley. Bradley's playing a position he's not used to playing and got handcuffed by a tough play. Why would any Dodgers fan want to boo him? He didn't try to commit the error. I vaguely understand booing a player who came into the season with a .300/.400/.500 line who's hitting .200/.240/.260 in August, but why boo a player who couldn't handle the ball on a weird play, especially if that player has a history of reacting poorly to negative reinforcement? I know, I know, this is kind of an ivory tower complaint, but this booing business just doesn't make sense to me.

Song of the Day: "Place To Be," Nick Drake

Monday, September 27, 2004

1994 Career Years

Doing research for a different project that should be ready some time this week, I noticed that Paul O'Neil was having a massive career year in 1994. When the strike aborted the season on August 12, 1994, O'Neil was hitting .359/.460/.603, good for a .354 EqA. O'Neil was a very good player for a long time, but that was O'Neil's only season with an EqA above .305. Moreover, Baseball Prospectus' rate stats indicate that O'Neil had his best season in the field, with a Rate2 of 110, his career high. O'Neil's season was good for 10.5 WARP3, which calculates wins contributed above a replacement player adjusted for season length and historical competition levels. But since BP only adjusts for (162/season-length)*2/3 (to account for skewing the totals for players from 19th century short seasons), that total is artificially low if used as a pro-rating comparison, so O'Neil arguably should be credited with 12.5 Wins Above Replacement using WARP3. His second-highest single-season WARP3 was 9.7. 12.5 is pretty darned good; Barry Bonds only topped that total three times before he became super-human in 2001, and twice his total was "only" 12.6.

What would Paul O'Neil's season look like if he'd kept up that level of play and the strike hadn't occurred? The Davenport Translations, which adjust for park and era, have O'Neil belting 34 home runs and 32 doubles, driving in 108 runs, and putting up a .369/.468/.641 line (good for a .360 EqA). How did O'Neil accomplish that? Well, the .369 batting average should be a bit of a tip-off that he perhaps had good luck with balls in play. He had a .376 batting average on balls in play (BABIP) in 1994, well above his career .306 mark. That was only part of it, though, as he had his third best AB/2B year, his second best AB/K year, his best PA/BB year, and his best AB/HR year by a longshot. He was probably helped by the small sample size and perhaps he would have slumped a bit in September as he typically had to that point in his career, but he almost certainly would have had his hands-down career year had the season continued.

Who else had career years in 1994 that were cut short? The one you probably know about is Matt Williams, who had a decent shot at breaking Roger Maris' then-record for home runs in a single season. Using the same adjustment as with Paul O'Neil, Williams' WARP3 should perhaps have been 10.8 when adjusted for full season-length.That would best his 1993 career best of 9.4, but that should be taken with a pinch of pepper since Williams had played 112 of SF's 115 games to that point and it probably would have been a struggle to keep him in the lineup that frequently without a decline in his productivity. If you're familiar with the WARP scale, that 10.8 figure probably strikes you as too low for someone on pace to tie Maris. Well, Williams wasn't much for getting on-base. His OBP in 1994 was .319, only slightly better than his career .317 OBP. Between his above average whiff rate and his below average walk rate, his actual value probably was less than his perceived value.

Frank Thomas was in the middle of an historic season when the strike began. He was hitting .353/.487/.729, all career highs, with a .379 EqA and 38 home runs. His full-season adjusted WARP3 amounts to 13.2, easily a career high.Offensively, this was Frank Thomas' best year, but defensively it was his career lowpoint, a pretty remarkable accomplishment for a fielder as unaccomplished as Thomas. His EqA adjusted for all time was .397, which tops every season of Ted Williams' career except his magical 1941, every season of Babe Ruth's career except his first with the Yankees, and every pre-2001 season of Barry Bonds' career (have you heard? Bonds has been pretty good lately). Thomas' season, like O'Neil's and Williams', was fueled by a career high in AB/HR. He also had his second best year for doubles power and for PA/BB. He also benefited from a higher than normal BABIP, .336 versus his career .311. Thomas probably wouldn't have finished the season with an era-adjusted .397 EqA had the strike not occurred, but it was nonetheless one of the five or so best offensive seasons by a player not named Bonds in my lifetime.

Jeff Bagwell's 1994 was only slight less historic than Thomas'. He had a .382 EqA (.381 adjusted for all time) while playing excellent defense at first base. He posted a career high in AB/HR and was enjoying a 2-year period in which he didn't strike out too frequently. His walk rate actually wasn't all that impressive and wasn't close to his level since. BABIP certainly aided his season, as his was .353 versus his career .317 mark.

Kenny Lofton easily had his best season in 1994, with a career high .315 EqA and one of his better years defensively. Lofton managed a career high in WARP1 (wins above replacement not adjusted for era) despite the short season. Lofton posted career highs with a .349 AVG, a .412 OBP and a .536 SLG (his next best year for SLG was at .453), as well as a career high in Isolated Power. In keeping with our implicit theme, Lofton had a .369 BABIP, well above his career .335 BABIP.

A few others come close to this dubiously-dubious distinction. You may have heard about the season this Tony Gwynn fellow had. Gwynn's .394/.454/.568 line constituted three career highs and his EqA was likewise a career high. This wasn't Gwynn's best year overall, though, as his defense at age 34 was lagging and his age 27 season in 1987 still takes the cake. Gwynn also benefited from the BABIP fairy, with a .388 versus his career .340. Albert Belle also had a monster year, posting a career-best EqA. But Belle's bad defense in 1994 makes it his second best year behind 1995. Ken Griffey, Jr. had probably his best offensive season, but his superior defense in 1991 and 1993 make those years more valuable than his 1994. Will Clark had an outstanding year, but accounting for the influence of The Ballpark in Arlington shows his 1988 and 1989 to be clearly superior. Mickey Tettleton had an excellent year, but his raw .248/.419/.463 line is inflated by the overall offensive increase in 1994, and his offense and defense were both superior in 1989 and 1991. Chili Davis had an excellent year, but it was pretty much matched by his 1995. Bob Hamelin quite clearly had a career year, but his career was too brief and mercurial for that to matter. Shane Mack had his best year for rate stats but only played 81 games and didn't contribute much defensive value. Kevin Mitchell had an amazing year, but it didn't match his 1989. Plenty of other players had career years without managing to have great years and thus didn't catch my eye, pero asi es la vida.

You remember listening to Paul Harvey on the radio and how he would always have that nice conclusion? "And ... now ... you know ... the rrrrrrrrest ... of the story." Anyway, I don't have one of those, so just go ahead and move right along.

Song of the Day: "The Sea Horse," Yo La Tengo

Sunday, September 26, 2004

Braving Bad Puns

After the Braves clinched the division, the MLB index had a section on it with the heading "Joy Division." After recovering from the cognitive dissonance created by an Ian Curtis-related baseball pun, I endeavored to address every baseball analyst's albatross and see just how those darned Georgians keep pumping out the banners.

You and I both know that I wasn't about to come up with some capital-A Answer, so instead I just focused on the way that the Braves roster has been put together. Here's a table of how many players of each acquisition type have been major contributors (arbitrarily defined as 150+ plate appearances or 40+ innings pitched) to the Braves in each of their division championship years (and we'll throw in their second-place finish in 1994). I split acquisitions into ten categories:
Once a player had been acquired, they remained the same acquisition type throughout their stint with the Braves, with the exception of free agents this year becoming previous free agent signings in subsequent years. That is, contract extensions and re-signings after a player had been granted free agency didn't change that a player was acquired by trade and so forth. Here are the resulting numbers:

Players 91 92 93 94 95 96 97 98 99 0 1 2 3 4
FA, that year 5 0 3 1 0 1 0 4 3 2 5 6 3 4
FA, prev 3 7 4 6 3 3 2 1 4 4 3 2 5 3
Trade, minors 2 2 2 1 1 1 1 1 1 0 1 2 2 2
Trade, <27 1 0 0 0 0 0 2 2 0 0 0 0 0 2
Trade, 27+ 2 5 3 3 2 3 5 3 7 7 6 4 5 5
HS 5 5 6 7 7 8 8 5 6 4 6 7 4 3
College, young 4 4 3 3 2 3 1 0 0 0 0 0 0
College, >20 0 0 0 0 2 1 1 0 0 0 1 1 2 4
Intl FA 0 0 0 1 1 2 3 3 6 3 4 4 4 2
MinL FA 0 0 1 1 1 1 0 2 0 1 1 1

Total 22 23 22 23 19 23 23 21 27 21 27 27 25 25

Here's how that breaks down into players who were developed in the minor leagues by the Braves and players who weren't:

Players 91 92 93 94 95 96 97 98 99 0 1 2 3 4
Non-AtlDevel 11 12 10 10 5 7 9 10 14 13 14 12 13 14
Atl Devel 11 11 12 13 14 16 14 11 13 8 13 15 12 11

Those numbers, however, don't tell the whole story, so I entered each player's contribution in wins above replacement using WARP1 to see how value was distributed among the acquisition types:

WARP1 91 92 93 94 95 96 97 98 99 0 1 2 3 4
FA, that year 15.9 0 14.6 2.3 0 3.4 0 12.4 5.5 2.7 3.6 7.4 4.1 8.2
FA, prev 11.5 23 9.5 15.1 13.6 9.9 11.9 10.9 10.1 19.2 23.5 11.8 10.3 9.6
Trade, min 5.2 12.6 9.2 3 5.5 10.6 8.5 6.4 7.4 0 2.4 6.1 8.1 5
Trade, 26- 2.9 0 0 0 0 0 4.3 2.7 0 0 0 0 0 8
Trade, 27+ 5.4 11.8 11.9 8.6 11.3 13.8 20.7 11.9 17.3 17 9.8 14 24.1 23
HS 28.3 19.6 27 18.8 29.1 32.8 28.1 27.1 35.5 23 23.4 27.3 18.9 11.6
College,young 11.6 14.3 16.1 9.4 7.3 5.6 7.5 0 0 0 0 0 0 0
College,20+ 0 0 0 0 2.4 1.3 1.3 0 0 0 1.8 2.4 2.8 6.6
Intl FA 0 0 0 1.6 3.9 7.1 11.8 19.4 16.8 18.6 14.9 21.6 24.1 13.3
MinL FA 0 0 4.5 2.4 3.5 3 0 4.8 0 2.7 2.2 2.6 0 0
Total 80.8 81.3 92.8 61.2 76.6 87.5 94.1 95.6 92.6 83.2 81.6 93.2 92.4 85.3

Players 91 92 93 94 95 96 97 98 99 0 1 2 3 4
NotAtlDevel 35.7 34.8 36 26 24.9 27.1 36.9 37.9 32.9 38.9 36.9 33.2 38.5 48.8
AtlDevel 45.1 46.5 56.8 35.2 51.7 60.4 57.2 57.7 59.7 44.3 44.7 60 53.9 36.5

Several trends stand out (or at least appear to stand out to somebody who spends hours doing the research). First, Atlanta obviously isn't too keen on drafting college players. In the early days, they had a few community college kids like Dave Justice and Jeff Blauser, but from 1998 to 2003 they had close to nothing in productivity from college-drafted players. Their high school picks have annually fueled the team, contributing the most wins of any segment of the team until 2003. In 2003 and 2004, international free agents have taken over that title, though most of that production is confined to Andruw Jones, Rafael Furcal, and Javy Lopez. Nearly all of the production from players acquired as minor leaguers in trades comes from John Smoltz, who himself was only two years out of high school, so 60% or so of the Braves' production during this stretch has come from players acquired before their 20th birthday. I don't have the data to compare that figure to other teams in general, so let's use this year's Dodgers for comparison. Of the Dodgers with 40+ IP or 150+ PA, only Adrian Beltre and Eric Gagne were acquired by LA before their 20th birthday, and despite those two being the Dodgers' two most valuable players they've only contributed 22% of LA's WARP1. Clearly, the Braves have done a tremendous job of harvesting very young talent.

The second clear trend is that the Braves haven't much for signing big-impact free agents. 38% of their WARP from free agents has come in the form of Greg Maddux. The rest have mainly been spare parts, with only Terry Pendleton and Brian Jordan being major exceptions (ok, we can add Andres Galaragga to that list too). Atlanta has, however, had pretty good success with some free agent reclamation pitching projects in recent years with the likes of John Burkett, John Thomson, and Jaret Wright. The Braves have tended towards acquiring more free agents in recent years, but much of that has to do with the dissipation of the team's young core. The Braves' key veteran acquisitions have tended to come in trades, including J.D. Drew, Mike Hampton, Russ Ortiz, Gary Sheffield, Kenny Lofton, Denny Neagle, Marquis Grissom, Fred McGriff, and Otis Nixon, few of whom stayed with the team for very long.

The third trend is more questionable. It appears that the Braves might finally have shifted to a roster constructed more from veteran acquisitions. Aside from the youth-and-strike-heavy 1994-1996 period, the WARP the Braves have gotten from veteran players has been between 32 and 39 each year until 2004, when it has taken off. On the other side, the WARP from Atlanta-developed players looks to be at its lowest outside of 1994 by a wide margin. Whether this is a change in organizational philosophy toward resurrecting salary-dumped veterans like Drew, Hampton, and Ortiz or whether this is simply a momentary blip before players like Andy Marte and Jose Capellan tip the scales back remains to be seen. It should be noted, however, that the Braves have gotten pretty good value in return for young pitchers they'd developed: Jason Schmidt netted Denny Neagle (whose 13.0 WARP in 2+ years nearly matches Schmidt's 14.9 for the Pirates over five years), Damian Moss netted Russ Ortiz, Odalis Perez and Andrew Brown (with Brian Jordan) netted Gary Sheffield, Tim Spooneybarger netted Mike Hampton, and Jason Marquis and company netted J.D. Drew and company. The Braves, then, are perhaps maximizing the value of their organizational resources by acknowledging Leo Mazzone's apparent ability to greatly improve any pitcher's performance and acting to accumulate position players and lower-risk pitchers.

John Schuerholtz, Bobby Cox (who also was the team's general manager for the five years preceding 1991 and helped rebuild the farm system), Paul Snyder, Chuck LaMar, and Leo Mazzone all deserve a ton of credit for the Braves success, and how much credit each individual deserves is unclear and probably beside the point. When folks argue for the approach of taking high risk/high ceiling high school prospects, their argument is certainly bolstered by the Braves' success. The Braves' strategy is probably not likely to succeed in scenarios where a team doesn't have talented baseball personnel and abundant resources, which is why it has been widely copied but never matched. One wonders, however, if some of the young "Moneyball"-type executives will begin to follow the Braves' path in situations where resources for intensive player evaluation and development are available. This approach might very well be the best way to maximize the value of organizational resources for larger market teams who now find themselves competing with more performance-analysis-oriented teams. Paul DePodesta and the Dodgers could and to some extent are following the same strategy. With a renowned scouting director in Logan White and a renowned pitching coach in Jim Colborn as well as a strong market, the Dodgers are probably as good a candidate as any to replicate the Braves' success, and their continued reliance on high school players in the 2004 amateur draft under DePodesta's watch is an indication that this might be the direction they're headed in.

Song of the Day: "Off Minor," Coltrane/Monk

This page is powered by Blogger. Isn't yours?